Quantification of four major metabolites of embryotoxic N -methyl- and N -ethyl-2-pyrrolidone in human urine by cooled-injection gas chromatography and isotope dilution mass spectrometry

Article abstract of DOI:10.1021/ac300439w

N-Methyl- and N-ethyl-2-pyrollidone (NMP and NEP) are frequently used industrial solvents and were shown to be embryotoxic in animal experiments. We developed a sensitive, specific, and robust analytical method based on cooled-injection (CIS) gas chromatography and isotope dilution mass spectrometry to analyze 5-hydroxy-N-ethyl-2-pyrrolidone (5-HNEP) and 2-hydroxy-N- ethylsuccinimide (2-HESI), two newly identified presumed metabolites of NEP, and their corresponding methyl counterparts (5-HNMP, 2-HMSI) in human urine. The urine was spiked with deuterium-labeled analogues of these metabolites. The analytes were separated from urinary matrix by solid-phase extraction and silylated prior to quantification. Validation of this method was carried out by using both, spiked pooled urine samples and urine samples from 56 individuals of the general population with no known occupational exposure to NMP and NEP. Interday and intraday imprecision was better than 8% for all metabolites, while the limits of detection were between 5 and 20 μg/L depending on the analyte. The high sensitivity of the method enables us to quantify NMP and NEP metabolites at current environmental exposures by human biomonitoring.

Full text of DOI:10.1021/ac300439w

Article
pubs.acs.org/ac
Quantification of Four Major Metabolites of Embryotoxic N-Methyl-
and N-Ethyl-2-pyrrolidone in Human Urine by Cooled-Injection Gas
Chromatography and Isotope Dilution Mass Spectrometry
,
†
†
†
‡
†
Birgit K. Schindler,* Stephan Koslitz, Swetlana Meier, Vladimir N. Belov, Holger M. Koch,
†
†
†
Tobias Weiss, Thomas Bru
̈
ning, and Heiko U. Kafferlein
̈
†
Institute for Prevention and Occupational Medicine of the German Social Accident Insurance, Ruhr-University Bochum (IPA),
rkle-de-la-Camp Platz 1, 44789 Bochum, Germany
Max Planck Institute for Biophysical Chemistry, Facility for Synthetic Chemistry, Am Fassberg 11, 37077 Go
Bu
̈
‡
̈
ttingen, Germany
*
S Supporting Information
ABSTRACT: N-Methyl- and N-ethyl-2-pyrollidone (NMP
and NEP) are frequently used industrial solvents and were
shown to be embryotoxic in animal experiments. We
developed a sensitive, specific, and robust analytical method
based on cooled-injection (CIS) gas chromatography and
isotope dilution mass spectrometry to analyze 5-hydroxy-N-
ethyl-2-pyrrolidone (5-HNEP) and 2-hydroxy-N-ethylsuccini-
mide (2-HESI), two newly identified presumed metabolites of
NEP, and their corresponding methyl counterparts (5-HNMP,
2
-HMSI) in human urine. The urine was spiked with deuterium-labeled analogues of these metabolites. The analytes were
separated from urinary matrix by solid-phase extraction and silylated prior to quantification. Validation of this method was carried
out by using both, spiked pooled urine samples and urine samples from 56 individuals of the general population with no known
occupational exposure to NMP and NEP. Interday and intraday imprecision was better than 8% for all metabolites, while the
limits of detection were between 5 and 20 μg/L depending on the analyte. The high sensitivity of the method enables us to
quantify NMP and NEP metabolites at current environmental exposures by human biomonitoring.
INTRODUCTION
Besides ingestion and inhalation, uptake of NMP can also
■
9
−11
occur by penetration through the skin.
In particular, dermal
N-Alkyl-2-pyrrolidones are organic solvents, miscible with
water and capable of dissolving nonpolar, polar, and even
inorganic chemicals. For this reason, they are essential
industrial solvents used worldwide in the production of
polymers, coating materials, paints, pesticide formulations,
absorption is an important route of uptake at the work-
1
2,13
place,
whereas ingestion has to be considered an additional
route of uptake in the general population. Human biomonitor-
ing can integrate exposure via all three routes and is therefore
1
2,13
1
,2
an important tool for risk assessment.
The determination of
cosmetics, and drugs. The two most important compounds of
this class are N-methyl-2-pyrrolidone (NMP) and N-ethyl-2-
pyrrolidone (NEP). The majority of data, however, is available
specific metabolites in urine and plasma is a well-established
14−16
technique for exposure assessment of NMP at workplaces.
NMP is successively oxidized to 5-hydroxy-N-methyl-2-
pyrrolidone (5-HNMP), N-methylsuccinimide (MSI), and 2-
hydroxy-N-methylsuccinimide (2-HMSI) in humans (Figure
3
for NMP, whereas studies related to NEP are scarce.
NMP is a high-production volume chemical and its annua2l
world production in 2003 was estimated to be up to 150 000 t.
1
). 5-HNMP and 2-HMSI are major metabolites of NMP in
4
,5
The acute toxicity of NMP is low. However, studies in rats
have shown dose-related decreases in fetal body weights after
oral and inhalation exposure and at exposure levels below
maternal toxicity. In addition, fetal malformations were found at
17
urine, accounting for 44% and 20% of the administered dose.
Conjugation of metabolites with glutathione or sulfation was
not observed.
Analytical methods for the determination of 5-HNMP or 2-
6,7
maternal toxic doses. For these reasons, NMP has recently
been classified in category 1b for developmental toxicity (may
cause harm to the unborn child) by the European Union thus
HMSI in urine have been previously reported with use of gas
17−20
chromatography and mass spectrometry (GC-MS),
liquid
chromatography and tandem mass spectrometry (LC-MS/
8
21,22
resulting in restrictions and limitations in its use. All industrial
MS),
and gas chromatography in combination with flame
formulations and consumer products containing ≥5% NMP
from now on must be labeled as embryotoxic worldwide
according to the Globally Harmonized System of Classification
and Labelling of Chemicals (GHS).
Received: February 14, 2012
Accepted: March 21, 2012
Published: March 21, 2012
©
2012 American Chemical Society
3787
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Article
1
7
Figure 1. Oxidative metabolism of NMP according to Åkesson et al. (A) and proposed metabolism of NEP based on the known metabolites of
NMP (B) (ox = oxidation).
2
3
ionization detection (GC-FID). Up to now, these methods
and 2-HESI metabolites and their deuterated analogues (5-
1
4,16
have been exclusively used to study work-related exposures
or experimental exposures of volunteers in inhalation
Internal standardization and quantification
were carried out by using commercially available 5-HNMP
HNEP-d , 2-HESI-d ). Then we developed and validated a
5
5
sensitive and specific analytical method based on solid phase
extraction of 5-HNMP, 2-HMSI, 5-HNEP, and 2-HESI from
urine samples, followed by silylation of the analytes. The
quantification was carried out by cooled-injection gas
chromatography and isotope dilution mass spectrometry
(GC-MS). Finally, we assessed environmental exposures to
NMP and NEP in persons of the general population, and thus
proved the practical suitability of our method in a field study.
2
4−26
chambers.
and 2-HMSI standards and their deuterium labeled ana-
logues5-hydroxy-N-methyl-2-pyrrolidone-d (5-HNMP-d4)
4
and 2-hydroxy-N-methylsuccinimide-d (2-HMSI-d )in the
3
3
case of analytical methods based on mass spectrometry. The
limits of detection (LODs) were between 6 and 1000 μg/L for
5
-HNMP and 30 and 1000 μg/L for 2-HMSI and therefore
suitable for the quantification of the internal dose after
occupational exposure to NMP.
In contrast to NMP, information on NEP is limited. The
acute oral toxicity of NEP has been reported to be
EXPERIMENTAL SECTION
Reagents and Materials. Acetonitrile (ACN), ethyl
acetate (EtOAc) and methanol (MeOH) for GC, ethanol
EtOH), and acetic acid (100%) of highest available analytical
grade were purchased from VWR International (Darmstadt,
Germany). N-(tert-Butyldimethylsilyl)-N-methyltrifluoroaceta-
mide (MTBSTFA) and pyridine were purchased from Sigma-
Aldrich (Seelze, Germany). SPE cartridges containing hydroxy-
■
(
5
approximately 4-fold higher than that of NMP. After oral
administration, NEP has been shown to display a similar
spectrum of developmental toxic and teratogenic effects in
rodents, as was reported previously for NMP. The onset of the
observed effects, however, was less pronounced. The use of
NEP is expected to gain importance due to the classification
and labeling requirements of NMP in the EU. In many
industrial formulations and consumer products NMP is already
substituted by the nonclassified NEP due to their very similar
27
+
lated polystyrene−divinylbenzene copolymer (Isolute ENV , 1
mL, 100 mg) were obtained from Biotage AB (Uppsala,
Sweden). Deionized water was prepared by a Milli-Q Plus
system (Millipore, Eschborn, Germany). Nitrogen for solvent
evaporation during sample preparation and helium as carrier
gas for GC was of 4.5 and 6.0 quality. 5-HNMP, 2-HMSI, and
their deuterated analogues 5-HNMP-d and 2-HMSI-d were
28
physicochemical properties.
Owing to the toxicity of NMP and NEP and their broad
application ranges, it is necessary to study both workplace and
low-level environmental exposures. Low levels of NMP have
been detected in the environment such as municipal and
4
3
purchased from Ramidus AB (Lund, Sweden) and were ≥95%
isotopic purity. Since analytical standards for 5-HNEP and 2-
HESI were not commercially available, they were synthesized
and characterized along with their deuterated analogues 5-
HNEP-d and 2-HESI-d . For this purpose, N-ethylsuccinimide
1
,29,30
industrial wastewaters and in US drinking water supplies.
Exposure of the general population to low environmental levels
of NMP and NEP has not been studied yet mostly due to the
lack of sensitive and specific analytical methods. Only the
5
5
(TCI Deutschland GmbH Eschborn, Germany), (±)-malic
acid, and a 70% aqueous solution of ethyl amine (Sigma-
Aldrich, Seelze, Germany), acetonitrile-d3 (99.8%, Deutero
GmbH, Kastellaun, Germany), and LiAlD4 (99%, Fluka
Analytical, Buchs, Germany) were used as starting materials.
Instrumentation for Standard Characterization. All
22
previously published method by Suzuki et al. is considered
sensitive enough to assess low-level exposures to NMP (LODs
of 6 and 30 μg/L for 5-HNMP and 2-HMSI). This method,
however, has not been used to assess exposure of NMP in the
general population. In addition, no analytical method at all for
the determination of NEP metabolites is available so far.
Based on the well-known metabolism of NMP in humans, we
postulated the formation of 5-hydroxy-N-ethyl-2-pyrrolidone
1
13
synthesized standards were characterized by H and C NMR,
1
MS, HPLC, and IR. H NMR was performed on a Varian
MERCURY-400-MR spectrometer at 400 MHz. Chemical
shifts (δ) are given in ppm relative to TMS. 13C NMR was
performed at 100 MHz on a Varian 400-MR spectrometer,
using attached proton test for multiplet detection. Multiplicities
of signals are described as singlet (s), doublet (d), triplet (t),
(
5-HNEP) and 2-hydroxy-N-ethylsuccinimide (2-HESI) as
major metabolites of NEP in humans (Figure 1). To prove
this hypothesis, we first synthesized and characterized 5-HNEP
3
788
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Article
quartet (q), and multiplet (m). Mass spectra of the synthesized
standards were obtained with electro spray ionization (ESI),
using LCQ and ESI-TOF mass spectrometers (MICROTOF
toluene. Toluene (500 mL) was added and the mixture was
refluxed for 24 h with a Dean−Stark trap for the azeotropic
removal of water. After cooling to RT, the reaction mixture was
washed with 80 mL of water and 50 mL of brine, then dried
over Na SO . The solvent was evaporated in vacuo, and the
(focus), Bruker, Bremen, Germany). A Knauer smartline
gradient HPLC system (Knauer, Berlin, Germany) including
2
4
two pumps (1000), UV detector (2500), column thermostat
residue was recrystallized from a hexane−ether mixture (5/1,
(
25 °C), and a 20 μL injection loop was used during standard
v/v) as a colorless solid (4.67 g, yield 72%) and with a mp of
31
preparation. A Eurospher-100 C column (5 μm, 250 × 4 mm,
81.5−82.4 °C (lit. mp 81−82 °C). Characterization: HPLC:
1
8
Knauer, Berlin, Germany) was used as the analytical column at
a flow rate of 1.1 mL/min. HPLC grade water containing 0.1%
trifluoroacetic acid (TFA) was used as solvent A and
acetonitrile containing 0.1% TFA as solvent B. The detection
of the analytes was carried out at 220 nm. Purity of the
standards was controlled by thin-layer chromatography on
Macherey Nagel ready-to-use plates with silica gel (G/UV254),
using 10% molybdenum−phosphoric acid in ethanol (Sigma-
Aldrich, Seelze, Germany) as a developer. Column chromatog-
raphy was carried out on MERCK silica gel (grade 60, 0.040−
0−3 min, 99% A, 3−13 min, 99−95% A, t = 9.6 min, area:
R
1
100%. H NMR (Figure S-4, Supporting Information,
CD OD): δ 1.12 (t, J = 7.2 Hz, 3 H, CH ), 2.44 (dd, A-part
3
3
of ABX-system, J
= 18.0 Hz, JAX = 4.4 Hz, 1 H,
AB
a
b
CH H CH OH), 3.01 (B-part of ABX-system, J = 17.9 Hz,
x
AB
a
b
x
JBX = 8.3 Hz, 1 H, CH H CH OH), 3.50 (q, J = 7.2 Hz, 2 H,
x
CH N), 4.53 (dd, J = 4.4 Hz, JBX = 8.3 Hz,1 H, CH OH)
2
AX
13
ppm. C NMR (Figures S-5 and S-6, Supporting Information,
CD OD): δ 13.4 (CH ), 34.7 (CH N), 39.1 (CH ), 68.2
3
3
2
2
(CHO), 176.8 (CO), 180.0 (CO) ppm. EI-MS (Figure S-
•
+
0.063 mm). IR spectra (KBr) were recorded on Spectrum BXII
7, Supporting Information): m/z (%) 143 (42%) [M] , 115
•
+
+
(
Perkin-Elmer, Waltham, USA).
(76%) [M − C H ] , 72 (100) [C H NO] . ESI-MS positive
2
4
3
6
+
Synthesis of Standards. (±)-5-Hydroxy-N-ethyl-2-pyr-
rolidone (5-HNEP). N-Ethylsuccinimide (500 mg, 3.93 mmol)
was dissolved in 26 mL of anhydrous ethanol. The solution was
mode: m/z (%) 166 (100) [M + Na] ; negative mode: m/z
−
−
(%) 285 (92) [2M − H] , 142 (100) [M − H] . IR (KBr): ν
3373 (OH), 2992/2977/2961/2940 (CH), 1685 (CO). Purity
was determined as >99%.
cooled in an ice bath, and 30 mg of NaBH was added in small
4
portions at regular intervals (10−15 min), followed by the
addition of 2 drops of 4 M HCl in dioxane after each portion of
Ethylamine-d (C D NH ). Ethylamine-d (C D NH ) was
5
2
5
2
5
2
5
2
32
prepared from acetonitrile-d and LiAlD according to Meese
3 4
NaBH . The course of the reaction was monitored by HPLC
with the following modification. After quenching the reaction
with water, volatile amine, ether, and water were slowly distilled
under gradually diminishing pressure (p = 500−50 mbar) into a
receiving flask, cooled with a mixture of dry ice and acetone. In
contrast to the original report, however, the flask did not
contain HCl. Cold ethanol was carefully added to the distillate,
until it became homogeneous (at 0 °C). The yield was
considered to be 80%, and the intermediate was not isolated.
Half of this solution was used for the preparation of N-
4
(
solvent A/B = 90/10, retention times (t ) of the starting
R
compound 7.6 min, product 4.0 min, side product 3.9 min).
The reaction was complete in about 2 h, when 300 mg (7.93
mmol) of NaBH and 20 drops (∼1 mL) of HCl were added.
4
The solvent was evaporated in vacuo (50−60 mbar) at 0 °C
into a trap cooled with an acetone−dry ice bath, and the
residue was extracted with 100 mL of cold CH Cl . The
2
2
suspension was filtered through a glass filter, and the filter cake
was washed with 2 × 25 mL of cold CH Cl . The combined
ethylsuccinimide-d and the other half for the synthesis of 2-
HESI-d5.
2
2
5
solutions were evaporated in vacuo at 0 °C. The residue was
taken up in CH Cl , transferred to a column with 75 g of SiO ,
(±)-N-Ethylsuccinimide-d . Succinic acid anhydride (4.0 g,
2
2
2
5
and eluted with a CH Cl /MeOH mixture (gradient 20:1 →
40 mmol) was added at 0 °C to a stirred solution of 2.8 g (56
2
2
10:1) to obtain 5-HNEP as a colorless oil (300 mg, yield 59%,
mmol) of ethylamine-d in an ether/ethanol/water mixture
5
R = 0.4 (CH Cl /MeOH, 15:1)). Characterization: HPLC: 0−
(10/2/1, v/v/v), and the reaction mixture was stirred at RT
overnight. Volatile materials were evaporated in vacuo and the
residue was coevaporated with 2 × 25 mL of toluene. Then,
200 mL of toluene was added and the mixture was refluxed for
24 h with a Dean−Stark trap for azeotropic removal of water.
After cooling to RT, the reaction mixture was washed with 50
mL of water and 25 mL of brine and dried over Na SO . The
f
2
2
3
min, 99% A; 3−13 min, 99−95% A; 13−20 min, 95−60% A;
1
t = 12.9 min, area: 97%. H NMR (Figure S-1, Supporting
R
Information, CD OD): δ 1.17 (t, J = 7 Hz, 3 H, CH ), 1.82 (m,
3
3
a
b
1
H, CH H CH(OH)), 2.29 (m, 2 H, CH CO), 2.49 (m, 1 H,
CH H CH(OH)), 3.20 (A-part of ABX system, J = 14.4 Hz,
JAX = 7.2 Hz, 2 H, CH H N), 3.43 (CH H N), 5.22 (dd, J = 6.4
and 2.51, 1 H, CHOH) ppm. C NMR (Figure S-2,
2
a
b
3
AB
a
b
a
b
2
4
1
3
solvent was evaporated in vacuo, and the residue was filtered
Supporting Information, CD OD): δ 12.8 (CH ), 28.6
through 25 g of SiO with a hexane/ethyl acetate mixture (1:1,
3
3
2
(
CH ), 29.7 (CH ), 35.5 (CH N), 83.8 (CHO), 176.5 (C
v/v) as an eluent. Evaporation of the solvents in vacuo yielded
2
2
2
O) ppm. EI-MS (Figure S-3, Supporting Information): m/z
(
N-ethylsuccinimide-d as oil (3.9 g, yield 74%). Character-
5
•
+
+
1
%) 129 (100%) [M] , 112 (45%) [M − OH] , 96 (30%) [M
ization: HPLC: solvents A/B = 90/10, t = 7.6 min, H NMR
R
+
•+
−
H O − CH ] , 85 (30) [C H NO] . ESI-MS, positive
(CDCl ): δ 2.57 (s, CH ).
2
3
4
7
3
2
+
+
mode: m/z (%) 281 (100) [2M + Na] , 152 (58) [M + Na] ;
(±)-5-Hydroxy-N-ethyl-2-pyrrolidone-d5 (5-HNEP-d ). 5-
5
−
negative mode: m/z (%) 257 (100) [2M − H] , 128 (69) [M
−
HNEP-d5 was prepared from N-ethylsuccinimide-d5 and
−
H] . IR (KBr): ν 3337 (OH), 2978/2939 (CH), 1670 (CO).
NaBH in anhydrous ethanol with 4 M HCl, as described
4
Purity was determined as >97%.
-Hydroxy-N-ethylsuccinimide (2-HESI). (±)-Malic acid
5.36 g, 40.0 mmol) was added at 0 °C to the stirred solution
of 4.5 mL of 70% aqueous ethylamine (d = 0.8, 2.5 g of EtNH2,
6 mmol) in 50 mL of ethanol and the reaction mixture was
above for 5-HNEP. Characterization: HPLC: solvents A/B =
2
90/10, tR = 4.0 min, area 100%. 1H NMR (Figure S-8,
a
b
(
Supporting Information, CD OD): δ 1.82 (m, 1 H, CH H CH-
3
a
b
(OH)), 2.3 (mc, 2 H, CH CO), 2.49 (m, 1 H, CH H CH-
2
1
3
5
(OH)), 5.22 (dd, J = 2.6 and 6.4 Hz 1 H, CHOH) ppm. C
stirred at RT overnight. Volatile materials were evaporated in
NMR (Figure S-9, Supporting Information, CD OD): δ 29.0
3
vacuo, and the residue was coevaporated with 2 × 25 mL of
(CH ), 30.1 (CH ), 84.2 (CHO), 176.9 (CO) ppm. EI-MS
2
2
3
789
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Figure S-3, Supporting Information): m/z (%) 117 (100%)
Article
(
+
•+
[
M − OH] , 85 (17) [C H NO] . ESI-MS: positive mode: m/
4
7
+ +
z (%) 291 (100) [2M + Na] , 157 (65) [M + Na] ; negative
−
−
mode: m/z (%) 267 (100) [2M − H] , 133 (79) [M − H] . IR
KBr): ν 3337 (OH), 2952 (CH), 2233 (CD), 1666 (CO).
Purity was determined >98%.
±)-2-Hydroxy-N-ethylsuccinimide-d (2-HESI-d ). 2-HESI-
(
(
5
5
d5 was prepared from 5.36 g (40.0 mmol) of (±)-malic acid and
the solution of 2.8 g (56 mmol) of ethylamine-d in an ether/
5
ethanol/water mixture according to the method described
above for 2-HESI (mp 82.3−83.4 °C). Characterization:
HPLC: 0−3 min, 99% A, 3−13 min, 99−95% A, t = 9.6
R
1
min, area: 100%. H NMR (Figure S-10, Supporting
Information, CD OD): δ 2.46 (dd, A-part of ABX-system, J
3
AB
a
b
=
18.0 Hz, JAX = 4.4 Hz, 1 H, CH H CH OH), 3.01 (dd, B-part
x
Figure 2. Deuterated internal standards of 5-HNMP, 2-HMSI, 5-
HNEP, and 2-HESI used for sample preparation and isotope dilution
mass spectrometry.
of ABX-system, J
= 17.9 Hz, JBX = 8.3 Hz, 1 H,
AB
a
b
x
CH H CH OH), 4.53 (dd, J = 4.4 Hz, JBX = 8.3 Hz,1 H,
AX
x
13
CH OH) ppm. C NMR (Figure S-11, Supporting Informa-
tion, CD OD): δ 39.1 (CH ), 68.3 (CHO), 176.9 (CO),
Sample Preparation. After equilibration to RT the urine
samples were homogenized and an aliquot of 600 μL was
transferred to a 2-mL vial. Then, 200 μL of the IS working
solution and 200 μL of 1% HAc in water were added. The
3
2
1
80.0 (CO) ppm. EI-MS (Figure S-7, Supporting
•
+
Information): m/z (%) 148 (48%) [M] , 120 (76%) [M −
•
+
+
C H ] , 77 (100) [C3D5HNO] . ESI-MS positive mode: m/z
2
4
+
+
(
[
%) 171 (100) [M + Na] ; negative mode: m/z (%) 295 (89)
2M − H] , 147 (100) [M − H] . IR (KBr): ν 3373 (OH),
664/2361/2237 (CD), 1685 (CO). Purity was determined as
solution was homogenized on a vortex mixer. The ENV solid
−
−
phase extraction cartridges were placed onto a Vacmaster SPE
station (Biotage AB, Uppsala) and preconditioned with 2 × 500
μL of EtOAc/EtOH (4/1, v/v), 2 × 500 μL of MeOH, and 4 ×
500 μL of 1% HAc in water. The preconditioned ENV+
cartridges were loaded with 1 mL of the diluted urine sample.
The cartridges were washed with 250 μL of 1% HAc in water
and 750 μL of water. Afterward the extraction cartridges were
dried in vacuo. The analytes were eluted with 1.75 mL of
EtOAc/EtOH (4:1, v/v) into 2-mL vials. The eluate was
evaporated to dryness in a stream of nitrogen. The residue was
reconditioned in 200 μL of acetonitrile and again evaporated to
dryness in order to remove trace residues of water prior to
derivatization. The residue was then dissolved in 30 μL of
pyridine and 30 μL of MTBSTFA was added. Derivatization
was performed at 110 °C within 60 min. After being cooled to
RT the sample was diluted with 50 μL of EtOAc and
transferred to a microinsert. Cooled injection GC-MS analysis
was performed by injection of 2 μL of each sample. If necessary,
the final extracts could be stored up to 2 weeks at −20 °C until
analysis.
2
>
99%.
Preparation of Standards and Quality Control Ma-
terial. Individual stock solutions (1 g/L) of 5-HNMP, 2-
HMSI, 5-HNEP, and 2-HESI were prepared by weighing 10 mg
of each compound in a 10-mL-volumetric flask and dissolving
with acetonitrile. One milliliter of each stock solution was
combined in a 10-mL-volumetric flask and diluted with water to
the mark yielding a working solution 1 of 100 mg/L for each
standard. Working solution 1 was diluted with water to a
concentration of 50 mg/L (working solution 2) and 5 mg/L
(
working solution 3). Six calibration standards (0.05 to 10 mg/
L) were prepared by diluting the working solutions described
earlier in pooled urine from multiple donors. The pooled urine
was frozen, thawed, and filtrated before use. An unspiked urine
sample from the same pooled urine was used as the urine blank
sample. The calibration standards and the urine blank sample
were aliquoted in 1 mL portions that could be stored at −20 °C
up to 1 year. Within each analytical run a full set of calibration
standards including a urine blank sample and a reagent blank
Cooled Injection Gas Chromatography/Isotope Dilu-
tion Mass Spectrometry. Cooled injection GC-MS analysis
was performed on an Agilent 7890 GC system with a CTC
CombiPAL autosampler (Agilent Technologies, Waldbronn,
Germany) and a KAS 4 cold injection system (Gerstel GmbH
(
water) was analyzed.
Individual stock solutions (1 g/L) of the deuterium-labeled
standards (5-HNMP-d , 2-HMSI-d , 5-HNEP-d , and 2-HESI-
4
3
5
d5, Figure 2) were prepared in acetonitrile as described above
for the stock solutions of the nonlabeled standards. An internal
standard IS working solution (5 mg/L of each standard) was
prepared in water by combining 50 μL of each stock solution in
a 10-mL-volumetric flask and diluting with water to the 10 mL
mark.
Certified quality control material is not available for
metabolites of NMP and NEP. Therefore we prepared in
house quality control (QC) material at two concentrations (0.2
and 2.0 mg/L). For this purpose, 0.2 and 2.0 mL of working
solution 1 (100 mg/L) each were diluted in a 100-mL-
volumetric flask, using pooled urine from multiple donors.
Again, the pooled urine was frozen, thawed, and filtrated prior
use. QC material of both concentrations was analyzed within
each analytical run. All solutions were stored at −20 °C in the
dark, making them stable up to 1 year.
̈
& CoKG, Muhlheim a.d. Ruhr, Germany), whereas mass
spectrometric detection was performed on an Agilent 5975
mass spectrometer. A DB-35MS capillary column (60 m, 0.25
mm i.d., 0.25 μm film thickness, Agilent Technologies,
Waldbronn, Germany) was used for chromatographic separa-
tion. Helium 6.0 (1 mL/min) was used as the carrier gas. The
KAS injector was operated in solvent vent mode (split 1:50) for
0.5 min after injection and splitless between 0.5 and 2.3 min.
The injector was programmed from 40 °C (0.5 min) to 240 °C
(120 deg/min; 0 min) and to 260 °C (600 deg/min; 10 min).
The column temperature was 50 °C (4 min), raised to 90 °C
(25 deg/min; 1 min), then to 190 °C (10 deg/min; 3 min) and
to a final temperature of 280 °C (30 deg/min; 10 min). The
injector was cooled to the initial temperature with air at the end
of the oven program. The temperature of the transfer line was
3
790
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Article
2
80 °C, the ion source was set to 230 °C, and the quadrupole
RESULTS AND DISCUSSION
■
was set to 150 °C. Electron ionization was carried out at 70 eV.
Selective and sensitive mass traces were used for quantitation
No data are currently available on the metabolism of NEP,
making it difficult to identify specific metabolites of NEP that
can be used for method development. Thus, we postulated the
formation of 5-HNEP and 2-HESI based on the previously
reported formation of 5-HNMP and 2-HMSI in humans after
(
quantifier) and confirmation (qualifier) of the target analytes
in single ion monitoring mode (Table 1). The ratio of
Table 1. Retention Times, Quantifiers and Qualifiers, and
Their Ratios of the Four Metabolites of NMP and NEP (tR,
retention time)
17
exposure to NMP. 5-HNEP and 2-HESI, however, were not
commercially available. Therefore, in-house synthesis of 5-
HNEP and 2-HESI along with their deuterated analogues was
necessary.
analyte
tR [min]
quantifier [m/z]
qualifier [m/z]
ratio
53
Standard Synthesis. Synthesis of 5-HNEP and other N-
alkyl derivatives has been previously reported in the course of a
multistep sequence via a spontaneous cyclization of the
theanine Strecker aldehyde and starting from cis-4-octen-1,8-
5
5
2
2
5
5
2
2
-HNMP
-HNMP-d4
-HMSI
21.45
21.41
21.27
21.23
21.69
21.65
21.40
21.34
172
176
186
189
186
191
200
205
98
102
144
147
112
39
61
31
-HMSI-d3
-HNEP
3
3,34
dioic acid.
The most straightforward synthetic approach to
-HNEP-d5
-HESI
5-HNEP, however, is based on the transformation of succinic
acid anhydride into N-ethylsuccinimide in the presence of
ethylamine, followed by a partial reduction with NaBH . The
158
163
-HESI-d5
4
reduction of cyclic imides with NaBH in the presence of HCl
4
35
quantifier and qualifier in the samples had to be consistent with
the ratio determined from standard solutions. Variations of
was shown to provide 5-hydroxy-N-alkyl-2-pyrrolidones. The
exact amount of HCl, however, was not specified and this
reaction has not been described for N-ethylsuccinimide and
≤
30% of the ratio quantifier to qualifier were accepted.
Method Validation. Method validation involved the
NaBH so far. We found that the outcome of this reaction
4
determination of the limit of detection (LOD), recovery,
intraday and interday imprecision, and matrix effects. The LOD
was defined as a signal-to-noise ratio of 3. For calibration, the
ratio of the peak area of each standard vs internal standard was
plotted against the respective concentration. The slope of the
calibration function was used for quantification because no
urine without traces of NMP and NEP metabolites was
available for calibration. Water was used as reagent blank during
method validation. Intraday and interday imprecision was
determined in pooled urine samples at two concentrations (0.2
and 2.0 mg/L). Relative standard deviations (RSD) were
determined for intraday and interday variations, based on a set
of 8 and 12 measurements. Recovery also was tested at 0.2 and
strongly depends on the amount of added HCl as well as on the
temperature during the reduction and the workup procedures.
The best results were obtained when 2 mol of NaBH and 1
4
mol of HCl were added sequentially in small portions to 1 mol
of imide in ethanol at 0 °C. An excess of HCl caused
“
overreduction” or led to instability of the product. The
removal of the solvent and extraction of 5-HNEP with
dichloromethane also had to be carried out at 0 °C. The
synthesis of 2-HESI was achieved by the cyclization of malic
acid with ethylamine.
In addition, the synthesis of stable isotope-labeled standards
of 5-HNEP and 2-HESI was necessary for internal stand-
ardization by isotope dilution. Commercially available standards
of 5-HNMP-d4 and 2-HMSI-d3 are labeled in the ring.
Therefore, we first tried to introduce the deuterium labels
into the pyrrole ring to obtain the corresponding ethyl
2.0 mg/L. For this purpose, eight different urine samples with a
wide range of creatinine concentrations between 0.26 and 3.11
g/L were selected to consider varying urine compositions when
determining the recovery of the analytes rather than using
pooled urine samples with a particular creatinine level.
The eight urine samples were analyzed unspiked to
determine the background level of NMP and NEP metabolites
and spiked with 0.2 and 2.0 mg/L. The analyte concentrations
in the spiked and unspiked urine samples were calculated from
the ratio of the area of the analyte and internal standard based
on an external calibration curve. Finally, the relative recovery
was calculated by subtraction of the base levels derived in the
unspiked samples from the corresponding concentrations
determined in the spiked samples.
derivatives 5-HNEP-d4 and 2-HESI-d . For this purpose,
3
succinic anhydride-d and malic acid-d were used as starting
4
3
materials. Heating these compounds with ethylamine in toluene
indeed provided the required imides. The deuterium−hydrogen
exchange that was observed under these conditions, however,
made this approach inappropriate. The CD CONC H -groups
2
2
5
are prone to protonation at the oxygen atom (with participation
of the COOH-residue), and the protonated tertiary amide
residue may easily lose a deuterium atom at the α-position to
the carbonyl group. Therefore, we have chosen to introduce the
deuterium labels into the ethyl moiety of 5-HNEP and 2-HESI.
Synthesis of 5-HNEP-d and 2-HESI-d followed the same
5
5
Biological Samples. Spot urine samples from 56
individuals (31 females, 25 males) with a median age of 40
years (range 18−64) not known to be exposed to NMP or NEP
were collected to study the urinary concentrations of 5-HNMP,
route as described above for the nonlabeled compounds except
that ethylamine-d was not commercially available. For this
5
reason, ethylamine-d was prepared from acetonitrile-d and
5
3
LiAlD and immediately used to synthesize 2-HESI-d and N-
4
5
2-HMSI, 5-HNEP, and 2-HESI in the general population and
ethylsuccinimide-d . N-Ethylsuccinimide-d was smoothly re-
5
5
to prove that our method was sensitive and specific enough to
determine background levels of NMP and NEP metabolites in
the general population. All samples were stored at −20 °C until
analysis was carried out. An approval of the ethics committee of
duced to the required 5-HNEP-d at high yields. No loss of
deuterium labels was observed under these conditions. Purity of
5
all standard substances (5-HNEP, 2-HESI, 5-HNEP-d , and 5-
5
HESI-d ) was >97%.
5
the Ruhr Universitat
collection of the urine samples.
̈
Bochum (No. 3867-10) is available for the
Method Development. Once 5-HNEP, 2-HESI, and their
deuterium-labeled analogues were synthesized, purified, and
3
791
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Article
characterized, we were able to develop a highly sensitive and
specific analytical method. We included 5-HNMP and 2-HMSI
and their deuterated internal standards to quantify exposure to
̈
NMP and NEP in one analytical run. Methods of Jonsson et
1
8
20
al. and Bader et al. were taken into account during method
development. We used GC-MS since enhanced separation
efficiency and the detection of specific mass fragments were
required due to the structural similarities of the metabolites of
NMP and NEP.
The high polarity of NMP and NEP metabolites and the
complexity of the urinary matrix presented a challenge for
method development. Efficient separation of the metabolites
from urine was achieved by solid-phase extraction. The SPE
material (ENV+) chosen, a cross-linked styrene−divinylben-
zene copolymer, combines polar and nonpolar retention
mechanisms and was previously shown to be highly suitable
36
for extraction of polar metabolites from a urinary matrix. To
ensure a protonation of the hydroxyl-functions of the
metabolites, we slightly acidified the urine samples prior to
SPE. The best results for the elution step were achieved by
using a mixture of ethyl acetate and ethanol. Ethanol, however,
has to be completely removed prior to the derivatization step
since even trace amounts will significantly reduce the
derivatization yield. Derivatization of the hydroxyl-groups was
performed by silylation because the extracts can be directly
injected into the GC-MS system without any further cleanup
steps thus speeding up sample preparation and minimizing
losses during sample cleanup. Derivatization with MTBSTFA
was superior to that with BSTFA (N,O-bis(trimethylsilyl)-
trifluoroacetamide) in view of better chromatographic separa-
tion of the structurally very similar metabolites. In addition,
fragmentation was more specific for MTBSTFA derivates
compared to those after BSTFA derivatization, since the loss of
the methyl moiety of the BSTFA derivates was identical to the
mass difference between NMP and NEP metabolites. Pyridine
serves as a solvent of the residue that remained after
evaporation of the sample to dryness and as a catalyst in the
silylation reaction. The use of pyridine resulted in reduced
amounts of MTBSTFA, and therefore decreased contamination
of the analytical system, reducing instrument maintenance
when running a large number of samples.
Cold-injection was chosen to increase the sample volume for
GC analysis in order to improve LODs. At the same time
solvent vent mode was used to compensate for the increased
amounts of the injected solvent and MTBSTFA, in order to
protect the analytical column. Suitable mass traces for
quantitation and confirmation of the analytes were identified
in full scan mode (m/z 50−300) by injection of the derivatized
standards in ethyl acetate during method development. The
fragments with the highest abundance were the dimethylsilyl
ether fragments of the pyrrolidones and succinimides, which
were chosen for quantitation (quantifiers) later on. The same
fragments were detected for the deuterium-labeled internal
standards (Table 1). All analytes were well separated from one
another despite their structural similarity (Figure 3). The
stereoisomers of the analytes were not separated from each
other.
Figure 3. Chromatogram of a native urine sample from a person
representing the general population not occupationally exposed to
NMP and NEP (quantifier and qualifier are in black and gray). The
concentrations of 5-HNMP, 2-HMSI, 5-HNEP, and 2-HESI are 129,
8
7, 143, and 235 μg/L, respectively.
22
et al. with LODs in a similar range for 5-HNMP and 2-HMSI.
The calibration curve was linear from 0.05 to 10 mg/L with
coefficients of correlation of 0.998 or higher for all analytes,
thus covering a broad range of potential environmental and
occupational exposures. The slopes of the calibration curves in
water and urine samples were similar, indicating that there is no
significant matrix effect. Nevertheless, calibration was carried
out in urine rather than water to increase long-term stability of
the calibration standards during storage, based on the buffering
capacity of urine. The formation of artifacts from the use of
deuterium-labeled internal standard solutions during sample
preparation and measurement was excluded since none of the
reagent blanks spiked with internal standards contained any
traces of unlabeled NMP and NEP metabolites. The
reproducibility of the method presented as relative standard
deviations of the measured concentrations in spiked urine
samples was excellent and <8% for all analytes and at all spiked
levels (Table 2).
Intraday imprecision (n = 8) at 0.2 and 2.0 mg/L was 2.1−
4.9% and 1.5−4.1% depending on the analyte. Interday
imprecision (n = 12) at the same concentrations was similar
and determined to be 4.4−7.6% and 2.3−6.4%.
Since urine is an extremely varying and complex matrix, it is
necessary to determine the robustness of the method in
Method Validation. Excellent LODs of 20 (5-HNMP), 5
2-HMSI), 15 (5-HNEP), and 5 μg/L (2-HESI) were achieved.
(
Three-fold LODs were considered as limits of quantification.
Previously published methods efficiently analyze NMP
metabolites alone and reported 10−100-fold higher
1
8,20−23,37
LODs
with the exception of the method by Suzuki
3
792
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Article
a
Table 2. Validation Data for the Determination of NMP and NEP Metabolites
mean rel recovery
n = 8)
intraday imprecision
(
range rel recovery (%)
(%) (n = 8)
interday imprecision (%) (n = 12)
analyte
LOD (μg/L)
0.2 μg/L
2.0 μg/L
0.2 μg/L
2.0 μg/L
0.2 μg/L
2.0 μg/L
0.2 μg/L
2.0 μg/L
5-HNMP
2-HMSI
5-HNEP
2-HESI
20.0
5.0
91
100
96
97
99
83−112
97−101
92−106
98−113
92−107
95−101
95−101
100−109
4.9
2.3
2.1
2.8
4.1
1.9
1.5
2.0
6.1
7.6
6.2
4.4
6.4
4.9
5.1
2.3
15.0
5.0
98
104
104
a
Intraday and interday imprecision is presented as relative standard deviation of the spiked concentrations.
different samples. Urinary creatinine can be used as an indicator
for dilution and different matrix load of urine samples. For this
purpose, eight different urine samples of nonexposed persons
with creatinine concentrations from 0.26 to 3.10 g/L were
selected to cover the complete range of common urines as
recommended by the World Health Organization. The urine
samples were analyzed unspiked and spiked at two concen-
tration levels (0.2 and 2.0 mg/L) to determine the relative
recovery of the method. Metabolite levels of 44−277 μg/L (5-
HNMP), 12−92 μg/L (2-HMSI), <LOD−198 μg/L (5-
HNEP), and <LOD−288 μg/L (2-HESI) were quantified in
the unspiked urine samples of the recovery experiments. The
mean relative recoveries for all analytes were between 91% and
maximum levels and 95th percentiles of NEP metabolites
suggest that individuals within the general population are
presently exposed to NEP at levels comparable to NMP.
CONCLUSION
■
38
Our newly developed method provides for the first time a tool
for the simultaneous and specific analysis of major metabolites
of the two most frequently used N-alkyl-2-pyrrolidones, NMP
and NEP. This method was demonstrated to be highly robust
and selective, and therefore applicable for routine analysis. In
addition, the high sensitivity allows for the quantification of
NMP metabolites and presumed NEP metabolites in the low
microgram per liter range, making the method suitable for
environmental exposure assessment with a biomonitoring
approach.
1
04% (range: 83−112%) at 0.2 mg/L and 97−104% (range:
92−109%) at 2.0 mg/L, thus proving the high robustness of the
analytical method even for extremely varying urine samples.
Method Application. We used our newly developed
method to verify whether or not 5-HNEP and 2-HESI are
human metabolites of NEP and finally assessed environmental
exposures to both NMP and NEP in persons of the general
population with no known occupational exposure. We were
able to show that 5-HNEP and 2-HESI are human metabolites
of NEP thus proving our initial hypothesis that NEP is
metabolized in analogy to the NMP in humans. In addition, we
show for the first time that parts of the general population are
exposed to both NMP and NEP (Figure 3). NMP metabolites
were quantified in >96% of urine samples of the general
population, whereas NEP metabolites were found in >32% of
the samples (Table 3). The high incidence of NMP positive
ASSOCIATED CONTENT
Supporting Information
Additional information as noted in the text. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
*
S
AUTHOR INFORMATION
Corresponding Author
E-mail: schindler@ipa-dguv.de. Phone: +49-234-3024592.
Fax: +49-234-3024505.
■
*
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors are grateful to Ju
Frauendorf with co-workers (Institut fu
Biomolekulare Chemie, Georg-August-Universitat
̈
rgen Bienert (MPI) and Dr. Holm
r Organische und
Gottingen)
Table 3. Results of the Biomonitoring Study in 56 Persons
from the General Population with No Known Occupational
Exposure
̈
̈
̈
for recording NMR and mass spectra.
>LOD
(%)
median
[μg/L]
95 percentile
[μg/L]
maximum
[μg/L]
analyte
REFERENCES
■
5
2
5
2
-HNMP
-HMSI
-HNEP
-HESI
96.4
98.2
32.1
46.4
69.5
63.5
337.0
200.4
642.6
238.4
620.0
256.2
769.3
310.8
(
1) International Programme on Chemical Safety N-Methylpyrroli-
done, Concise International Chemical Assessment, Document 35;
Geneva, 2001; p 39.
2) OECD 1-Methyl-2-pyrrolidone, SIDS initial assessment report;
<15.0
<5.0
(
Paris, France, 2007.
samples in the general population reflects the ubiquitousness of
NMP in our environment. The lower prevalence of NEP
metabolites in urine may reflect the currently less frequent use
of NEP. Nevertheless, our results already indicate an increased
use of NEP as a substitute of NMP. This interpretation may
also be supported by the fact that higher median levels of NMP
metabolites (5-HNMP: 70 μg/L; 2-HMSI: 64 μg/L) were
found in urine compared to NEP metabolites (5-HNEP and 2-
HESI: <LOD), although up to now, no information is available
on kinetics and the quantitative fractions of 5-HNEP and 2-
HESI in human NEP metabolism. Nevertheless, the observed
(3) France CLH report Proposal for harmonized classification and
labelling of a chemical substance, Substance Name: N-ethyl-2-
pyrrolidone (NEP); 2011; pp 1−47.
(4) Bartsch, W.; Sponer, G.; Dietmann, K.; Fuchs, G. Arzneim. Forsch.
1
(
976, 26, 1581−1583.
5) Ansell, J. M.; Fowler, J. Food Chem. Toxicol. 1988, 26, 475−9.
(6) Saillenfait, A.; Gallissot, F.; Morel, G. Food Chem. Toxicol. 2003,
4
(
1, 583−8.
́ ́
7) Saillenfait, A. M.; Gallissot, F.; Langonne, I.; Sabate, J. P. Food
Chem. Toxicol. 2002, 40, 1705−12.
(
8) The Commission of the European Communities EU
Commission Regulation (EC) No 790/2009; 2009; Vol. 1, pp 235/
1−235/439.
3
793
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794

Analytical Chemistry
Article
(
9) Midgley, I.; Hood, a J.; Chasseaud, L. F.; Brindley, C. J.;
Baughman, S.; Allan, G. Food Chem. Toxicol. 1992, 30, 57−64.
10) Ursina, C.; Hansena, C. M.; Dykb, J. W. V.; Jensenc, P. O.;
Christensenc, I. J.; Ebbehoejd, J. Am. Ind. Hyg. Assoc. J. 1995, 7, 651−
60.
11) Payan, J.-P.; Boudry, I.; Beydon, D.; Fabry, J.-P.; Grandclaude,
M.-C.; Ferrari, E.; Andre, J.-C. Drug Metab. Dispos. 2003, 31, 659−669.
12) Bader, M.; Keener, S.; Wrbitzky, R. Int. Arch. Occup. Environ.
Health 2005, 78, 673−6.
13) Keener, S.; Wrbitzky, R.; Bader, M. Int. Arch. Occup. Environ.
Health 2007, 80, 327−34.
14) Bader, M.; Rosenberger, W.; Rebe, T.; Keener, S.; Brock, T. H.;
Hemmerling, H.-J.; Wrbitzky, R. Int. Arch. Occup. Environ. Health
006, 79, 357−64.
15) Nishimura, S.; Yasui, H.; Miyauchi, H.; Kikuchi, Y.; Kondo, N.;
(
6
(
(
(
(
2
(
Takebayashi, T.; Tanaka, S.; Mikoshiba, Y.; Omae, K.; Nomiyama, T.
Ind. Health 2009, 47, 355−62.
(
16) Anundi, H.; Langworth, S.; Johanson, G.; Lind, M.-L.; Åkesson,
̈
B.; Friis, L.; Itkes, N.; Soderman, E.; Jonsson, B. a. G.; Edling, C. Int.
Arch. Occup. Environ. Health 2000, 73, 561−569.
(
(
(
17) Akesson, B.; Jo
18) Jonsson, B. G.; Akesson, B. J. Chromatogr., B 1997, 694, 351−7.
19) Ligocka, D.; Lison, D.; Haufroid, V. J. Chromatogr., B 2002, 778,
̈
nsson, B. Pharmacology 1997, 25, 267−269.
̈
2
23−30.
20) Bader, M.; Rosenberger, W.; Will, W. In MAK collection for
Occupational Health and Safety. Part VI: Biomonitoring Methods; Vol.
1; Angerer, J., Schaller, K.-H., Eds.; Wiley-VCH: Weinheim,
Germany, 2008; pp 97−114.
21) Carnerup, M.; Akesson, B.; Jo
61, 107−13.
22) Suzuki, Y.; Endo, Y.; Ogawa, M.; Yamamoto, S.; Takeuchi, A.;
Nakagawa, T.; Onda, N. J. Chromatogr., B 2009, 877, 3743−7.
23) Kubota, R.; Endo, Y.; Takeuchi, A.; Inoue, Y.; Ogata, H.; Ogawa,
M.; Nakagawa, T.; Onda, N.; Endo, G. J. Chromatogr., B 2007, 854,
04−10.
24) Bader, M.; Wrbitzky, R.; Blaszkewicz, M.; Schap
Thriel, C. Arch. Toxicol. 2008, 82, 13−20.
25) Akesson, B.; Paulsson, K. Occup. Environ. Med. 1997, 54, 236−
0.
26) Carnerup, M. A.; Spanne, M.; Jo
62, 139−45.
27) Saillenfait, A.; Gallissot, F.; Sabate,
(
1
(
̈
nsson, B. J. Chromatogr., B 2001,
7
(
(
2
(
̈
er, M.; van
(
4
(
̈
nsson, B. Toxicol. Lett. 2006,
1
(
́
J. P.; National, I.; Recherche,
D.; Bourgogne, A. D. J. Appl. Toxicol. 2007, 491−497.
(
28) BASF corporation BASF at the European Coatings show in
Nuremberg, news release http://www.basf.com/group/
pressemitteilungen/P-09-163 (accessed Oct 27, 2011).
(
29) Samuel, V. L. GC-MS Analysis of Organics in Drinking Water
Concentrates and Advanced Waste Treatment Concentrates, Vol. 3;
Research Laboratory: Research Triangle Park, NC, 1985.
(
2
(
1
(
1
(
2
30) Matsui, S.; Okawa, Y.; Ota, R. Water Sci. Technol. 1989, 20,
01−210.
31) Lukes, R.; Pergal, M. Collect. Czech. Chem. Commun. 1962, 27,
387−1392.
32) Meese, C. O. J. Labelled Compd. Radiopharm. 1983, 20, 1047−
059.
33) Tanaka, T.; Watarumi, S.; Fujieda, M.; Kouno, I. Food Chem.
005, 93, 81−87.
34) Barco, A.; Benett, S.; Pollini, G. P.; Baraldi, P. G.; Simoni, D.;
(
Vicentini, C. B. Synthesis 1979, 68−70.
(
(
35) Hubert, J. C. Tetrahedron 1975, 31, 1437−1441.
36) Schindler, B. K.; Forster, K.; Angerer, J. J. Chromatogr., B 2009,
̈
8
77, 375−381.
37) Lokajova,
J. Chem. 2011, 9, 825−833.
38) WHO (World Health Organization) In Biological monitoring of
chemical exposure in the workplace; 1996; pp 20−51.
(
́
J.; Porras, S. P.; Elovaara, E.; Wiedmer, S. K. Cent. Eur.
(
3
794
dx.doi.org/10.1021/ac300439w | Anal. Chem. 2012, 84, 3787−3794